GB2329781A

GB2329781A - Motion picture compression

Info

Publication number: GB2329781A
Application number: GB9720488A
Authority: GB
Inventors: Kang-Young Lee; Hee-Jong Lee
Original assignee: Samsung Electronics Co Ltd
Current assignee: Samsung Electronics Co Ltd
Priority date: 1996-03-11
Filing date: 1997-09-26
Publication date: 1999-03-31
Anticipated expiration: 2017-09-26
Also published as: DE19743221C2; GB2329781B; DE19743221A1; GB9720488D0

Abstract

A motion picture compression circuit includes a first frame memory 20 for storing the start frame of motion picture information constructing a picture; second and third frame memories 10 for storing two successive frames of the motion picture information; a picture boundary detector 30 for detecting a picture boundary from the motion picture information of the two successive frames supplied from the second and third frame memories 10; a motion vector detector 40 for predicting a motion vector from the motion picture information of the start frame and end frame; a controller 50 for generating data required for forming a picture, according to the detected picture boundary and motion vector; and a picture reproducer 60 for reproducing an omitted frame using the data generated by the controller, and inserting it between the start frame and end frame.

Description

MOTION PICTURE COMPRESSION CIRCUIT AND METHOD USING MOTION VECTOR The present invention relates to a circuit and method for compressing motion picture information in a video signal processing system, specifically, to a motion picture compression circuit and method using motion vectors.

To view motion pictures of a film or TV, 30-50 frames/sec of information is needed. Thus, there is little time variation between successive frames in a movie film or videotape. In the digital image processing, data compression is carried out to reduce the redundancy between the successive frames.

The most general method of compressing video data is spatial information compression in which one frame is partitioned into 8x8 pixels of blocks, and discrete fourier transform (DCT) is performed for them. Here, the DCT is carried out for luminance and chrominance signals.

Another method is time information compression in which a motion which occurs between frames is predicted and compensated. The motion prediction is, generally, made for the luminance signal in the 16x16 pixels of block.

These information compression methods are executed independently, or combined to produce higher compression effect. The information compression is done by predicting motions between 2-4 successive frames, so that there is a difference in the amount of information according to the characteristic of frame. However, information about the entire frame is needed for the data compression.

In order to perform play, fast forward or rewind for the compressed video information, random access may be carried out by the group of picture (GOP), which is suggested by the MPEG. The GOP groups by about 0.5 seconds (approximately fifteen pictures). As an example, video information compression according to MPEG-I is explained below. Figure 1 shows an example of arrangement of frames in GOP. Referring to Figure 1, fifteen frames (pictures) form one GOP. The GOP has an I frame (Intracoded) picture which is necessarily included for the purpose of independence of GOP, a P frame (Predictivecoded picture) which is forward-direction predictive-coded picture between frames, and a B frame (Bidirectionally predictive-coded picture).

Figures 2A and 2B show frame types and coding order.

The I frame is intra-coded, and the I frame and P frame are coded in an order identical to that of the original image. The B frame is coded in a manner that the I and P frames are processed first, and then the B frame is inserted between them. In other words, the I and P frames are coded in the order of al-- > a2-- > a3-- > a4 as shown in Figure 2A, and the B frame is coded in the order of bl- > b4-- > b2-- > b3 as shown in Figure 2B. As described above, the video information compression cannot be made without information about the entire frame constructing the video information. This is inefficient, for example, in a monotonous picture where the background is hardly changed but a portion varies.

An aim of embodiments of the present invention is to provide a high-efficiency compression circuit and method for motion picture information having large redundancy, which reproduces the original motion picture with several frames even if there is no information about all the frames constructing the picture.

According to an aspect of the present invention, there is provided a motion picture compression circuit which receives motion picture information by the frame and performs compression/encoding of it, the motion picture compression circuit including: a first frame memory for storing the start frame of motion picture information constructing a picture; second and third frame memories for storing two successive frames of the motion picture information; a picture boundary detector for detecting a picture boundary from the motion picture information of the two successive frames supplied from the second and third frame memories; a motion vector detector for predicting a motion vector from the motion picture information of the start frame and end frame; a controller for generating data required for forming a picture, according to the detected picture boundary and motion vector; and a picture reproducer for reproducing an omitted frame using the data generated by the controller, and inserting it between the start frame and end frame.

Preferably, the picture boundary detector comprises: means for obtaining the average of square of luminance components of corresponding frames by the pixel; and means for comparing the average with a predetermined threshold, and deciding corresponding frame as a picture boundary when the average exceeds the threshold.

Preferably, the motion vector detector comprises: means for receiving pixel blocks in the predetermined unit, and obtaining the difference between them, the pixel blocks constructing the start frame and end frame of a picture; and means for comparing the difference with a predetermined threshold, and detecting a pixel block above the threshold as a motion block; and means for detecting a motion vector, according to the location of the motion block.

The picture reproducer may comprise: means for obtaining the number of frame to be inserted between the start frame and end frame; means for obtaining the movement distance from the motion vector by the pixel; and means for obtaining the location of the motion block, according to the movement distance, the number of frame to be inserted, and corresponding frame number.

According to a second aspect of the present invention there is provided a method for encoding compressed motion picture information, comprising the steps of: dividing motion picture information applied by the frame into the start frame, end frame, and a frame adjacent to the end frame, and storing them; detecting a picture boundary from the motion picture information of the neighbouring two frames; predicting a motion vector from the motion picture information of the start frame and end frame; and generating data required for constructing each picture according to the detected picture boundary, predicted motion vector, and the number of frame to be inserted between the start frame and end frame, reproducing an omitted frame using the generated data, and inserting the reproduced frame between the start frame and the end frame.

For a better understanding of the invention, and to show how embodiments of the same may be carried into effect, reference will now be made, by way of example, to the accompanying diagrammatic drawings, in which: Figure 1 shows an example of arrangement of frames in GOP; Figures 2A and 2B show frame types and coding order; Figures 3A and 3B are imaginary diagrams of pictures showing that a plane moves from the lower left corner to the upper right corner; Figure 4 is a block diagram of a compression motion picture encoding device according to an embodiment of the present invention; and Figure 5 shows the form of a motion vector according to the movement of motion blocks.

With reference to the attached drawings, a preferred embodiment of the present invention is described below in detail. It should be noted that component reference numbers are consistent between all the drawing. While, there appear many particular details like components of a circuit written in the following description. These details are provided only for a general understanding of the present invention. It would be obvious to a person having general knowledge of this technology that the present invention can be embodied without these particular details. In describing the present invention, a detailed description will be omitted if a detailed description of the prior function or construction makes the point of the present invention ambiguous.

Figures 3A and 3B are imaginary diagrams of pictures showing that a plane moves from the lower left corner to the upper right corner. Figure 3A shows the start picture in which the plane is located at position A at time tl, and Figure 3B shows the final picture in which the plane is located at position B at time t2. Meanwhile, the reproduction of a motion picture is performed in such a manner that a continuous motion is divided into many stopped pictures, and the stopped pictures are rapidly projected. By doing so, the stopped pictures appear to a viewer as if they were smoothly continued pictures according to residue effect. As shown in Figures 3A and 3B, when the picture transmission is made with intermediate pictures omitted, its receiver can estimate how the plane moves from position A to position B according to an embodiment of the present invention.

Figure 4 is a block diagram of a compression motion picture information encoding device using motion vectors, and Figure 5 shows the form of a motion vector according to the movement of motion blocks. The compression of motion picture information and encoding process of the compression motion picture information using the motion vectors are described below in detail with reference to Figures 4 and 5.

An input video signal VI is stored in a frame memory 10 having capacity available for storing video information of one frame. The initial frame of frames constructing one motion picture, start frame f", is stored in a separate place, that is f' frame memory 20. A picture boundary detector 30 compares luminance signals of two continuous frames by the pixel, and obtains the average of squares of the luminance signals. Then, picture boundary detector 30 compares the obtained average value with a threshold which was experimentally determined previously. From the comparison result, when the average is higher than the threshold, it is considered that there is a picture boundary between the two frames.

In addition, when yrn(x,y) is regarded as luminance component value (Y) at position (x,y) of the mth frame, picture boundary detector 30 outputs a result as described in the following formula (1).

- AyOe C,CY"(x,y) - Y5-1(x,y)] (1) When the result exceeds the threshold, corresponding frame is decided as the picture boundary, and frame number n is transmitted to a controller 50, to be used when a picture reproducer 60 performs encoding, which will be explained below.

A motion vector detector 40 receives motion picture information of the start frame and end frame, as shown in Figures 3A and 3B, among the frames constructing the picture, and predicts the trace of a moving object between the two frames. In this embodiment, data about the two frames is input by the 16x16 pixels of block, and the luminance components between corresponding blocks are compared. Here, a portion where there is no motion has very small luminance component value, whereas a portion where there is a motion has large value. Thus, blocks including the moving object can be seized. This information is also transmitted to controller 50.

That is, the motion picture information is obtained from start frame fS and end frame fE of each picture. these two frames are stored in two frame memories 10 and 20. Motion vector detector 40 obtains the difference between two inputs supplied from the two frame memories 10 and 20, each of which is constructed in the unit of 16x16 pixels of block, and compares the difference with the threshold, to select a block which exceeds the threshold.

This block is called motion block. A motion vector MV as shown in Figure 5 is obtained using the location of the motion block for the start frame and end frame.

Picture reproducer 60 reproduces the original picture using the number of frame inserted between the start frame and end frame supplied from controller 50, and motion vector indicating the transition of motion. The number of frame to be inserted is obtained from start frame fS and end frame fE as described in following formula (2).

The number of frame to be inserted = fS - fE - 1 (2) Then, the movement distance is obtained by the pixel, from the motion vector, as described in the following formulae (3) and (4). x component movement distance = XE - XS ---(3) y component movement distance = yE - yS ---(4) Therefore, the location of motion block for the nth frame to be inserted equals the following formulae (5) and (6). x;x' + [(XE-Xs)/(fE-fs-l)]*n ---(5) y,'y' + [(yE~yS)/(fE~fS,1)] ---(6) While there are many other methods for processing start frame fS and end frame fE, the compression using DCT is the most general method.

As described above, for motion picture information having relatively simple background and large redundancy, the present invention can reproduce the original picture using information about the start frame and end frame of one picture, the number of frame to be inserted between the start and end frames, and motion vector, in contrast to the conventional method in which the compression is carried out for all the frames constructing the motion picture. Accordingly, compression efficiency of information can be maximized.

Therefore, it should be understood that the present invention is not limited to the particular embodiment disclosed herein as the best mode contemplated for carrying out the present invention, but rather that the present invention is not limited to the specific embodiments described in this specification except as defined in the appended claims.

The reader's attention is directed to all papers and documents which are filed concurrently with or previous to this specification in connection with this application and which are open to public inspection with this specification, and the contents of all such papers and documents are incorporated herein by reference.

All of the features disclosed in this specification (including any accompanying claims, abstract and drawings), and/or all of the steps of any method or process so disclosed, may be combined in any combination, except combinations where at least some of such features and/or steps are mutually exclusive.

Each feature disclosed in this specification (including any accompanying claims, abstract and drawings), may be replaced by alternative features serving the same, equivalent or similar purpose, unless expressly stated otherwise. Thus, unless expressly stated otherwise, each feature disclosed is one example only of a generic series of equivalent or similar features.

The invention is not restricted to the details of the foregoing embodiment(s). The invention extends to any novel one, or any novel combination, of the features disclosed in this specification (including any accompanying claims, abstract and drawings), or to any novel one, or any novel combination, of the steps of any method or process so disclosed.

Claims

CLAIMS 1. A motion picture compression circuit which receives motion picture information by the frame and performs compression/encoding of it, the motion picture compression circuit comprising: a first frame memory for storing the start frame of motion picture information constructing a picture; second and third frame memories for storing two successive frames of the motion picture information; a picture boundary detector for detecting a picture boundary from the motion picture information of the two successive frames supplied from the second and third frame memories; a motion vector detector for predicting a motion vector from the motion picture information of the start frame and end frame; a controller for generating data required for forming a picture, according to the detected picture boundary and motion vector; and a picture reproducer for reproducing an omitted frame using the data generated by the controller, and inserting it between the start frame and end frame.
2. The motion picture compression circuit as claimed in claim 1, wherein the picture boundary detector comprises: means for obtaining the average of squares of luminance components of corresponding frames by the pixel; and means for comparing the average with a predetermined threshold, and deciding corresponding frame as a picture boundary when the average exceeds the threshold.
3. The motion picture compression circuit as claimed in claim 1 or 2, wherein the motion vector detector comprises: means for receiving pixel blocks in the predetermined unit, and obtaining the difference between them, the pixel blocks constructing the start frame and end frame of a picture; means for comparing the difference with a predetermined threshold, and detecting a pixel block above the threshold as a motion block; and means for detecting a motion vector, according to the location of the motion block.
4. The motion picture compression circuit as claimed in claim 1, 2 or 3, wherein the picture reproducer comprises: means for obtaining the number of frame to be inserted between the start frame and end frame; means for obtaining the movement distance from the motion vector by the pixel; and means for obtaining the location of the motion block, according to the movement distance, the number of frame to be inserted, and corresponding frame number.
5. A motion picture compression circuit substantially as herein described with reference to the accompanying Figures.
6. A method for encoding compressed motion picture information, comprising the steps of: dividing motion picture information applied by the frame into the start frame, end frame, and a frame adjacent to the end frame, and storing them; detecting a picture boundary from the motion picture information of the neighbouring two frames; predicting a motion vector from the motion picture information of the start frame and end frame; and generating data required for constructing each picture according to the detected picture boundary, predicted motion vector, and the number of frame to be inserted between the start frame and end frame, reproducing an omitted frame using the generated data, and inserting the reproduced frame between the start frame and the end frame.
7. A method for encoding compressed motion picture information substantially as herein described with reference to the accompanying Figures.

7. A method for encoding compressed motion picture information substantially as herein described with reference to the accompanying Figures.

Amendments to the claims have been filed as follows 1. A motion picture compression circuit which receives motion picture information by the frame and performs compression/encoding of it, the motion picture compression circuit comprising: a first frame memory for storing the start frame of motion picture information constructing a picture; second and third frame memories for storing two successive frames of the motion picture information; a picture boundary detector for detecting a picture boundary from the motion picture information of the two successive frames supplied from the second and third frame memories; a motion vector detector for predicting a motion vector from the motion picture information of the start frame and end frame; a controller for generating data required for forming a picture, according to the detected picture boundary and motion vector; and a picture reproducer for reproducing an omitted frame using the data generated by the controller, and inserting it between the start frame and end frame.

2. The motion picture compression circuit as claimed in claim 1, wherein the picture boundary detector comprises: means for obtaining the average of squares of luminance components of corresponding frames by the pixel; and means for comparing the average with a predetermined threshold, and deciding corresponding frame as a picture boundary when the average exceeds the threshold.

3. The motion picture compression circuit as claimed in claim 1 or 2, wherein the motion vector detector comprises: means for receiving pixel blocks in the predetermined unit, and obtaining the difference between them, the pixel blocks constructing the start frame and end frame of a picture; means for comparing the difference with a predetermined threshold, and detecting a pixel block above the threshold as a motion block; and means for detecting a motion vector, according to the location of the motion block.

4. The motion picture compression circuit as claimed in claim 1, 2 or 3, wherein the picture reproducer comprises: means for obtaining the number of frame to be inserted between the start frame and end frame; means for obtaining the movement distance from the motion vector by the pixel; and means for obtaining the location of the motion block, according to the movement distance, the number of frame to be inserted, and corresponding frame number.

5. A motion picture compression circuit substantially as herein described with reference to the accompanying Figures.

6. A method for encoding compressed motion picture information, comprising the steps of: dividing motion picture information applied by the frame into the start frame, end frame, and a frame adjacent to the end frame, and storing them; detecting a picture boundary from the motion picture information of the neighbouring two frames; predicting a motion vector from the motion picture information of the start frame and end frame; and generating data required for constructing each picture according to the detected picture boundary, predicted motion vector, and the number of frame to be inserted between the start frame and end frame, reproducing an omitted frame using the generated data, and inserting the reproduced frame between the start frame and the end frame.